Urinary expression of let-7c cluster as non-invasive tool to assess the risk of disease progression in patients with high grade non-muscle invasive bladder Cancer: a pilot study

Among all bladder tumors, 70–75% are non-muscle-invasive (NMIBC: 70% Ta, 20% T1, 10% Tis) at first evaluation. NMIBC is a heterogeneous group composed of stages Ta (non-invasive papillary carcinoma), T1 (sub-epithelial connective tissue invasive carcinoma) and CIS (Tis Carcinoma in situ), and various grades including well (G1) moderately (G2) and poor (G3) differentiated tumors. Many large cohorts studies analyzed the long-term survival of NMIBC patients in general after transurethral resection (TUR) and intravesical chemotherapy [1]. Survival range varies between studies but are mostly higher than in muscle-invasive bladder cancer (MIBC), high grade (HG) Ta/T1 disease represents about 5 to 23% of NMIBC at first diagnosis, with a steady increase in incidence. It is considered as a highly malignant tumor with a variable and unpredictable biology as well as a critical aspect of management. Most of NMIBC cases are treated by TUR of the tumor, but about 50–70% undergoes recurrence within 2 years and 10–30% progresses to muscle-invasive disease [1]. The most accepted predictors of progression in these patients are tumor size, multifocality, presence of CIS and vascular invasion [2]. According to the European Organization for Research and Treatment of Cancer (EORTC) classification, any T1, HG/G3, CIS, recurrent multifocal and larger than 3 cm Ta G1-G2/Low grade (LG), are considered bladder cancer (BC) at high risk of progression to muscle-invasive disease [1, 3]. However, these features may not accurately reflect clinical outcome and only partially address a reliable risk-adjusted treatment planning. In the era of precision medicine, many efforts have been made to increase knowledge about disease biology, to find reliable biomarkers associated with progression in BC [3].

MicroRNA (miRNAs), short noncoding RNAs regulators of gene expression [4], are also present in several body fluids such as serum, plasma and urine, with a high degree of stability indicating their extensive potential as biomarkers [5, 6]. Recent literature has highlighted the dysregulation of several miRNAs, among them let-7c cluster, in BC tissue as well as in urine and blood, suggesting their potential role as diagnostic and prognostic biomarkers in BC patients [7‐12].

Let-7c belongs to the let-7 miRNA family that plays an important role in several cellular processes, and it is widely viewed as a tumor suppressor miRNA [13, 14]. Let-7c coding sequence, organized in a cluster with miR-99a and miR-125b, is localized on chromosome 21 and resides within intron 6 of the poorly characterized and recently named LINC00478 gene [15]. Consistent with its tumor suppressor activity, let-7c is downregulated in many cancer, including BC, and has been related to tumor progression [16‐18].

Pathological and clinical features of BC, such as the tumor stage/grade, metastasis, recurrence and chemo-sensitivity are well known predictors of cancer specific survival [3]. Since HG-NMIBC patients display a non-negligible risk of recurrence and progression, BCG bladder instillation (induction and maintenance) is recommended due to its proven efficacy on reducing both recurrence and progression rates [3]. A recent EORTC study confirmed the role and the optimal schedule of BCG treatment following TUR [1]. Up to now, given the high risk of disease progression in BCG-failure patients, RC has been the gold standard treatment for these patients [3]. Early detection of patients with NMIBC at higher risk of progression by a non-invasive molecular marker, avoiding unsuccessful BCG treatment, and early identification of patients who may really benefit from an early RC, are of critical importance especially with regards to precision medicine.

Analysis of miRNAs seems to be the most promising strategy for the development of non-invasive early diagnostic/prognostic biomarkers because they are highly stable within urine, are more resistant against nuclease degradation and require little handling care [5‐7].

Let-7c miRNA is widely accepted as a tumor suppressor miRNA involved in the regulation of oncogenic pathways and downregulated in many types of tumor tissue, including BC [13‐18]. In line with let-7c tumor-suppressor role, our analysis of the cluster expression levels in HG-NMIBC patient’s tissue showed that let-7c, as well as miR-99a and miR-125b, were highly downregulated compared to normal tissue samples. In urine, let-7c, as well as the other let-7c cluster members, miR-99a and miR-125b has been found down-regulated in BC versus HC [11, 22, 23]. In the current study, we investigated the potential value of urinary let-7c cluster as non-invasive diagnostic/prognostic biomarker of HG-NMIBC patients. Our data showed a significant upregulation of urinary let-7c cluster in BC patients’ samples compared to HC. This observation is in contrast with previous findings showing decreased levels of urinary let-7c cluster in BC patients’ samples [11, 22, 23].

Potential reasons for divergences with previous studies may resides in the several challenges related to studying circulating miRNAs, especially the urinary ones. Among these a high degree of inter-individual variability of extracellular miRNAs levels [6], and technical sources of variation between studies (e.g., the RNA extraction method, the normalization, the technological platforms used etc.). For these reasons, to avoid technical biases, we maintained constant sample processing through the whole study. Moreover, it has to be considered that the debate about the relative contribution of organs and/or tissue to miRNA in biological fluids is still open since detailed knowledge of molecular mechanisms governing miRNA release from normal and tumor tissues is still lacking [6, 24, 25]. However, several authors in previous reports have shown opposite miRNA expression profile in bio-fluids versus matched tumour tissues [26, 27]. Our results showing that urinary let-7c expression is significantly upregulated and, coherently with its role of tumour suppressor, down-regulated in matched tissue samples compared to normal controls, fit with these cases. Moreover, we investigated the let-7c cluster levels in the extracellular fraction of BC cells characterized by the phenotype of HG-NMIBC.

It has been hypothesized that extracellular miRNAs may be released through both active and passive mechanisms [24]. In vitro studies also proved that miRNAs are exported into the cell supernatant [27]. Considering this, we wondered whether BC cells, with different grades and invasive power, could secrete let-7c cluster miRNAs in the cell medium, and examined its levels originating directly from the tumor. Interestingly, we found that let-7c cluster is detected in the culture medium (extracellular fraction) of BC cells with HG-NMIBC phenotype. Of note, by comparing the extra- to intra-cellular ratio between let-7c cluster levels in different grades and invasive power BC cells, we found let-7c ratio significantly higher in HG-NMIBC compared the other cell phenotypes. Accordingly, a previous study reported an enrichment of let-7 miRNA family in the culture medium of gastric cells in relation with different tumorigenic and invasiveness propensities [27]. These data suggest that the increase of extracellular let-7c in HG-NMIBC cells could be a marker of disease at higher risk of progression to muscle-invasive BC. Moreover, the increased release of the tumor-suppressor let-7c in the culture medium found in the HG- compared to the LG-NMIBC cells could suggest a reduction of anti-cancer effect of let-7c in the cells, to maintain their oncogenic potential, as already observed in other tumor systems [27].

Several reports associate miRNA expression to cancer stage, also in BC [8, 11, 28]. In the present study we found a relevant association of increased urinary let-7c cluster levels in T1 stage disease. These data, even if provided by histological examination from TUR, suggests that the increased urinary let-7c cluster levels could early discriminate patients at higher risk of tumor progression as T1 HG-NMIBC. In matched tumour tissues, the T1 versus Ta tumour stages comparison did not lead to substantial statistically differences, thus, only the urinary miRNAs expression shows the capability of discriminating the higher and more advanced T1 stage with a greater risk of tumor progression in HG-NMIBC patients from the Ta stage. These data, in liquid biopsy vs tissue specimens, suggest the importance of let-7c cluster in view of the long-term goal to potentially reduce or avoid all forms of invasive procedures. Regarding the prognostic stratification of NMIBC patients, EORTC tables, based on the six most significant clinical and pathological features, are well-known tools to calculate the risk of disease progression in NMIBC patients [29]. Risk stratification by EORTC progression score was confirmed in many reports. May et al. reported, in a series of 521 RC performed for MIBC, that baseline high risk NMIBC had significantly worse cancer survival after RC when compared with baseline low-to-intermediate EORTC progression risk score patients [30]. In our study, upregulation of urinary let-7c cluster has an impact on PFS probabilities thus suggesting that its expression may improve prognosis estimation, identifying patients at risk of progression. EORTC score and let-7c cluster expression under the threshold value, determined in ROC curve analysis, were significant predictors of PFS on univariable Cox regression analysis. On multivariable analysis, miRNA profiling was the only independent predictor of PFS in HG-NMIBC patients. In fact, in both intermediate and high risk EORTC score cohorts, patients with urinary expression for at least one member of the cluster displayed worse PFS (both log rank p < 0.001). On DCA, the benefit of using this prognostic model was evident in patients displaying a PFS > 20%. This cohort essentially includes all cases where the clinical decision making process between conservative options (intravesical BCG) and RC is controversial, being those patients with a PFS probability < 20% with a significant risk of disease progression, where the effectiveness of conservative treatments should be carefully balanced with the significant risk of progression.

The limitations of our study are the small sample size, and the tumor heterogeneity and multifocality that may hide variability in miRNA expression. Therefore, external validation of these findings with larger prospective studies is mandatory to confirm findings and to consider the clinical application of miRNA assays for diagnostic and prognostic purposes in HG-NMIBC.